3 ms·

In BigQuery I've found array_agg() to be preferable in most use cases. Also, in Hive I used to use Custom Reducers for many things where now I have to make do

In BigQuery I've found array_agg() to be preferable in most use cases. Also, in Hive I used to use Custom Reducers for many things where now I have to make do with Window Functions. BigQuery is blazingly fast though so it seems a reasonable sacrifice.

adwww 7y ago
I use pretty simple window functions with almost every BigQuery query I write, I can't imagine going back to a database that doesn't support them now.
- alexgmcm 7y ago
  Yeah - I've never had to use a db that didn't support them. It's almost unimaginable for me!